[ 343 ] 
creafed uniformly, when the fpecific gravities or ab- 
folute weights increafed uniformly. We fee from 
this, what conclufion fhould have been formed, had 
the increments of fpecific gravity from equal portions 
r . . w w + p 
of fait been equal. Again, fuppole that - , 
— L - b — , &c. denote a feries of fractions, whofe 
numerators, exprefiing the weights of a given quan- 
tity of water as increased by the addition of fait, and 
whofe denominators, exprefiing the bulks, both in- 
creafe uniformly, then will the feveral differences be- 
tween the 2d and ift, between the 3d and 2d, 
and fo on, be as m x m + q* m + qXm+xq* m-\- iq X 7 W+ 3J* 
=-■■■ — — — -7 &c. which fra&ions being inverfely 
as their denominators conftitute a decreafing feries ; 
but the increments of fpecific gravity from the 
addition of equal portions of fait, are proportionable 
to thefe fractions, and therefore ought perpetually to 
decreafe, though we allowed the bulk of the com- 
pound to be precifely equal to the bulk of the water 
and fait taken together, that is, though we allowed 
the bulk of the water to increafe uniformly accord- 
ing to the quantity of fait added : now as it is 
evident from Dr. Lewis’s experiments, and from 
each of the preceding tables, that the increments of 
fpecific gravity do decreafe upon the whole, when 
the abfolute weights increafe uniformly, we may 
venture to conclude that the bulks increafe uniformly 
alfo. I thought proper to explain the foregoing 
principle and to determine the ratio, becaufe the 
matter 
